{"id":"https://openalex.org/W4396650940","doi":"https://doi.org/10.48550/arxiv.2405.00955","title":"Recovering Labels from Local Updates in Federated Learning","display_name":"Recovering Labels from Local Updates in Federated Learning","publication_year":2024,"publication_date":"2024-05-02","ids":{"openalex":"https://openalex.org/W4396650940","doi":"https://doi.org/10.48550/arxiv.2405.00955"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2405.00955","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2405.00955","pdf_url":"https://arxiv.org/pdf/2405.00955","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2405.00955","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5008845106","display_name":"Huancheng Chen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Huancheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5067602750","display_name":"Haris Vikalo","orcid":"https://orcid.org/0000-0002-7945-4114","observed_orcids":["https://orcid.org/0000-0002-7945-4114"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Vikalo, Haris","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9994999766349792,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9994999766349792,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.0005000000237487257,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12357","display_name":"Digital Media Forensic Detection","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/gradient-inversion-attacks","display_name":"gradient inversion attacks","score":0.8840000033378601},{"id":"https://openalex.org/keywords/label-reconstruction","display_name":"label reconstruction","score":0.5600000023841858},{"id":"https://openalex.org/keywords/federated-learning-privacy","display_name":"federated learning privacy","score":0.925000011920929},{"id":"https://openalex.org/keywords/label-extraction","display_name":"label extraction","score":0.5299999713897705},{"id":"https://openalex.org/keywords/data-reconstruction","display_name":"data reconstruction","score":0.44600000977516174},{"id":"https://openalex.org/keywords/least-squares-estimation","display_name":"least squares estimation","score":0.7110000252723694},{"id":"https://openalex.org/keywords/data-heterogeneity","display_name":"data heterogeneity","score":0.8820000290870667}],"concepts":[{"id":"https://openalex.org/C2992525071","wikidata":"https://www.wikidata.org/wiki/Q50818671","display_name":"Federated learning","level":2,"score":0.5890527963638306},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5614916682243347},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.4019002914428711},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.34795743227005005}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2405.00955","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2405.00955","pdf_url":"https://arxiv.org/pdf/2405.00955","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2405.00955","doi":"https://doi.org/10.48550/arxiv.2405.00955","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2405.00955","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2405.00955","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2405.00955","pdf_url":"https://arxiv.org/pdf/2405.00955","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4396650940.pdf","grobid_xml":"https://content.openalex.org/works/W4396650940.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W3204999899","https://openalex.org/W4415707805","https://openalex.org/W7126107537","https://openalex.org/W4392822568","https://openalex.org/W4401508055","https://openalex.org/W4386766697","https://openalex.org/W7061377045","https://openalex.org/W6947833354","https://openalex.org/W7007933171","https://openalex.org/W7168257846"],"abstract_inverted_index":{"Gradient":[0],"inversion":[1],"(GI)":[2],"attacks":[3,194],"present":[4,71],"a":[5,72,134,153],"threat":[6],"to":[7,18,35,123],"the":[8,22,44,105,140,143,148,157,161,177,189],"privacy":[9],"of":[10,21,31,43,142,147,160,188,197],"clients":[11,106],"in":[12,47,63,100,107,152,192,195],"federated":[13],"learning":[14],"(FL)":[15],"by":[16,39,132],"aiming":[17],"enable":[19],"reconstruction":[20],"clients'":[23],"data":[24,37,149,172],"from":[25,79,139],"communicated":[26],"model":[27],"updates.":[28],"A":[29],"number":[30],"such":[32],"techniques":[33],"attempts":[34],"accelerate":[36],"recovery":[38,75],"first":[40],"reconstructing":[41],"labels":[42,131,146],"samples":[45],"used":[46,151],"local":[48,113],"training.":[49],"However,":[50],"existing":[51,182],"label":[52,74],"extraction":[53],"methods":[54],"make":[55],"strong":[56],"assumptions":[57],"that":[58,137,176],"typically":[59],"do":[60],"not":[61],"hold":[62],"realistic":[64,101],"FL":[65,109],"settings.":[66],"In":[67],"this":[68],"paper":[69],"we":[70],"novel":[73],"scheme,":[76],"Recovering":[77],"Labels":[78],"Local":[80],"Updates":[81],"(RLU),":[82],"which":[83],"provides":[84],"near-perfect":[85],"accuracy":[86],"when":[87],"attacking":[88],"untrained":[89],"(most":[90],"vulnerable)":[91],"models.":[92],"More":[93],"significantly,":[94],"RLU":[95,129],"achieves":[96],"high":[97],"performance":[98],"even":[99],"real-world":[102],"settings":[103],"where":[104],"an":[108],"system":[110],"run":[111],"multiple":[112],"epochs,":[114],"train":[115],"on":[116,167],"heterogeneous":[117],"data,":[118],"and":[119,156,171,184,200],"deploy":[120],"various":[121],"optimizers":[122],"minimize":[124],"different":[125],"objective":[126],"functions.":[127],"Specifically,":[128],"estimates":[130],"solving":[133],"least-square":[135],"problem":[136],"emerges":[138],"analysis":[141],"correlation":[144],"between":[145],"points":[150],"training":[154],"round":[155],"resulting":[158],"update":[159],"output":[162],"layer.":[163],"The":[164],"experimental":[165],"results":[166],"several":[168],"datasets,":[169],"architectures,":[170],"heterogeneity":[173],"scenarios":[174],"demonstrate":[175],"proposed":[178],"method":[179],"consistently":[180],"outperforms":[181],"baselines,":[183],"helps":[185],"improve":[186],"quality":[187],"reconstructed":[190],"images":[191],"GI":[193],"terms":[196],"both":[198],"PSNR":[199],"LPIPS.":[201]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2024,"cited_by_count":1}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2024-05-05T00:00:00"}